Explaining market value of players using SHAP

I am going to explain market predictions for two players Messi and Courtois. I still use fifa-23 dataset, but this time I base my prediction on one more variable - 'Club Position' to get better results. Models that I use are Random Forest Regressor and Catboost Regressor.

First model that I use is RandomForestRegressor. I have already used and shortly described in homework 1.

Second model is Catboost Regressor. The main idea of boosting is to sequentially combine many weak models and thus through greedy search create a strong competitive predictive model.

MLP - multi-layer perceptron is a neural network that consists of at least 3 layers. One layer usually consist of a linear and an activation.

4.

As we can see these two observations have different variables of the highest importance for RandomForestRegressor. Messi - 'Reactions', Courtois - 'Dribbling Total'.

5.

Variable 'Reactions' has positive influence on Messi's market value and negative for Courtois's market value.

6.

They differ mostly by small relatively number, but there are some cases when some variables are in top 4 for one package and are not in another.

7.

As we can see there are differences on variables importance on same observation between two presented models. For example: